Where the terms came from
AEO predates the LLM era — the term was coined around 2017–2018 to describe optimisation for Google's Answer Box and voice assistants like Alexa and Google Assistant. It re-emerged in 2023–2024 as those surfaces evolved into AI Overviews and conversational AI.
GEO was popularised by a 2023 academic paper (Aggarwal et al., 'GEO: Generative Engine Optimization') and adopted by SEO practitioners through 2024–2025. It explicitly addresses the generative nature of LLM outputs — the response is synthesised on the fly from multiple sources, not pulled verbatim from one ranked page.
Practical overlap and the small differences
About 80% of the work is identical: structured content, clear authorship, schema markup, ranking in conventional search, and earning brand mentions. AEO tends to emphasise tactics that win Featured Snippets and Answer Boxes (definitional first paragraphs, schema FAQ markup). GEO tends to emphasise tactics that survive synthesis (citation-worthy data points, comparative tables, original research the LLM has reason to reference).
Most reputable agencies treat them as one discipline with two emphases. If a vendor sells AEO and GEO as separate services with separate price tags, that's usually a packaging exercise rather than a methodological distinction.